Recovering the Pre-Fine-Tuning Weights of Generative Models.
Eliahu HorwitzJonathan KahanaYedid HoshenPublished in: CoRR (2024)
Keyphrases
- generative model
- fine tuning
- probabilistic model
- mixture model
- discriminative models
- conditional random fields
- prior knowledge
- linear combination
- object categories
- em algorithm
- discriminative learning
- semi supervised
- fine tuned
- hierarchical hidden markov models
- topic models
- data sets
- expectation maximization
- maximum entropy principle
- learning experience
- semi supervised learning
- naive bayes models
- generative and discriminative models
- decision trees